Spreads, depths, and quote clustering on the nyse and Nasdaq: Evidence after the 1997 sec rule changes Kee H. Chung,a,* Bonnie F. Van Ness,b Robert A. Van Nessb

10    ^During our sample time period, the TAQ database has data on 3,442 NYSE stocks. We exclude preferred stocks, warrants, and lower-class common stocks (e.g., class B and C common stocks) from the study sample. This leaves us a final sample of 2,912 stocks. There are over 22 million trades and 42 million quotes in our study sample.

11    ^When we replicate our analyses using other cutoff points for the composite match score, the results are qualitatively similar to those presented here.

12    ^A large number of quote updates for NYSE-listed stocks originate from off the NYSE. As Blume and Goldstein (1997) show, however, quotes that originate from off the NYSE only occasionally better NYSE quotes. Hence, we use only NYSE quotes in our study.

13    ^We estimate D_it using the algorithm in Lee and Ready (1991). We use quotes that are at least 15 seconds old. We acknowledge that effective spreads for Nasdaq-listed stocks may be underestimated due to inter-dealer trades.

14 ^When we compare the average depth of our Nasdaq sample of stocks calculated from the TAQ database with the corresponding figure calculated from the market maker quote data, we find that the former is significantly (about 50%) less than the latter.

15 ^We show only the plots for our NYSE stocks. We find similar results for our Nasdaq sample.

16 ^The results are available from the authors upon request.

17    ^See Bessembinder and Kaufman (1997) for a similar finding.

18    ^The stock price clustering was first noted in Harris (1991).

19    ^See, for example, Doran, Lehn, and Shastri (1995), Furbush (1995), Kleidon and Willig (1995), Laux (1995), Godek (1996), Huang and Stoll (1996), and Grossman et al. (1997).

20    ^See, for example, Christie, Harris, and Schultz (1994), Bessembinder (1997), Christie and Schultz (1999), and Barclay et al. (1999).

21    ^Sherwood Securities was the first defendant to settle litigation by agreeing to pay a total of $9.2 million on April 10, 1997. Kidder Peabody agreed to an out-of-court settlement of $14.1 million, followed by Herzog, Heine, Geduld Inc., who settled for $30.6 million on June 8, 1997. A global settlement ($900 million) that included all but one of the remaining 31 defendants was reached on December 24, 1997. The sum of all payments, including interest, totaled $1.027 billion.

22    ^See Ball, Torous, and Tschoegl (1985) and Grossman et al. (1997) for similar reasoning.

23    ^Grossman et al. (1997) suggest that it is human nature to prefer even numbers.

24    ^It is not clear whether these postulates are a realistic representation of the preference structure of liquidity providers. Ultimately, the reasonableness of these postulates can only be judged by the prescriptive power of their implications. That is, ultimate verdict on our model must depend on how well our model can explain empirical data.

25    ^Harris (1991) also noted similar empirical regularities in closing prices reported in the CRSP Daily Stock Master Database.

26    ^The frequency distributions of Nasdaq and NYSE quotes are significantly different from the uniform distribution that would be expected if quotes were randomly selected from the discrete set of sixteenths. We find that ²(15) = 729,416 for the Nasdaq sample and ²(15) = 216,836 for the NYSE sample.

27    ^Grossman et al. (1997) claim that the primary reason for lower clustering on the NYSE is the nature of the quote-generating process: NYSE quotes are created by a combination of public limit orders and specialist quotes. They argue that both investors and the specialist have strong incentive to place limit orders at odd-eighths as well as even-eighths. Since Nasdaq quotes also reflect limit order interest for our sample of stocks, the explanation given by Grossman et al. does not apply to our results.

28    ^See, for example, Christie and Schultz (1994), Godek (1996), and Barclay (1997).

29    ^See Harris (1991) for this definition of quote clustering.

30    ^It can be easily shown that the proportion of quotes in whole numbers equals QC + 1/16.

31    ^This result contradicts the findings of Huang and Stoll (1996) that after controlling for differences in economic factors, no relationship exists between quoted spreads and the frequency of odd-eighth quotes among their sample of 66 paired NYSE-Nasdaq stocks. However, our result is consistent with the findings of Barclay (1997), Bessembinder (1997), and Kandel and Marx (1997). These studies show that the spread is positively related to the degree of quote clustering.

32    ^Using the same procedure, it can be shown that 31% of the difference in the realized spread can be attributed to the differential quote clustering between the two markets.

33    ^Other likely sources of the wider spreads observed on the Nasdaq are commissions and the existence of alternative dealer quote dissemination systems. See Huang and Stoll (1996) for a detailed discussion of these issues.

34    ^See also Harris (1991) and Grossman et al. (1997) for similar reasoning.